Academic Editor: Pradeep Paudel
Background: The association between blunted dopaminergic neurotransmission and alcohol use disorder (AUD) is well-known. In particular, the impairment of postsynaptic dopamine 2 and 3 receptors (DRD2/3) in the ventral and dorsal striatum during the development and maintenance of alcohol addiction has been investigated in several positron emission tomography (PET) studies. However, it is unclear whether these changes are the result of adaptation or genetic predisposition. Methods: Here we investigated the association between DRD2/ankyrin repeat and kinase domain-containing 1 (ANKK1) TaqIA allele (rs1800497) status and striatal DRD2/3 availability measured by 18F-fallypride PET in 12 AUD patients and 17 sex-matched healthy controls. Age and smoking status were included as covariates. Results: Contrary to our expectations, TaqIA allele status was not associated with striatal DRD2/3 availability in either group and there was no significant difference between groups, possibly due to the relatively small sample size (N = 29). Conclusions: Nonetheless, this is the first in vivo study investigating the relationship between dopamine receptor availability and genetic factors in AUD. The pitfalls of assessing such relationships in a relatively small sample are discussed. Clinical Trial Registration: The published analysis is an additional, post hoc analysis to the preregistered trial with clinical trial number NCT01679145 available on https://clinical-trials.gov/ct2/show/NCT01679145.
Alcohol use disorder (AUD) is a complex and chronic disorder with high costs for society. The pathogenesis of AUD involves an interplay of social, individual, and biological factors; approximately 50% of the etiology of alcohol dependence is attributed to genetic influence [1].
Chronic alcohol intake is associated with lower dopamine receptor availability. In particular, reduced availability of dopamine D2 and D3 receptors (DRD2 and DRD3, respectively) in the striatum of patients with alcohol dependence compared to healthy controls (HCs) has been demonstrated by several positron emission tomography (PET) studies [2, 3, 4, 5, 6, 7, 8, 9] and described in recent reviews [10, 11]. However, it is unclear whether these changes reflect neuroadaptations to excess acute striatal dopamine release during regular substance use or are the result of genetic predisposition for substance use disorder. Several potential contributing genetic factors have been discussed in the literature. The Taq1A polymorphism of the dopamine D2 receptor DRD2 gene is a SNP (rs1800497) giving rise to the DRD2 allele (A2, rs1800497 (C)). The minor allele (A1, rs1800497 (T)) has been associated with a reduction of dopamine receptor availability and therefore been suggested to contribute to the development of substance use disorders also other mental disorders [12, 13, 14, 15].
More precisely, the minor allele (A1, rs1800497 (T)) has been linked to alcohol dependence in several studies [16, 17, 18, 19] although the effect size was found to be small [20].
Inconsistencies in published results have cast doubt on the functional relevance of the DRD2/ANKK1 TaqIA allele in AUD, as there is limited evidence that rs1800497 affects DRD2 role. A recent meta-analysis of 62 studies including 16294 participants found that the positive association between DRD2 and alcoholism was due to low allele frequencies in control subjects rather than changes in DRD2 gene expression [21]. On the other hand, single nucleotide polymorphism (SNP) of the DRD2/ANKK1 (rs1800497) TaqI allele was shown to be linked to reduced striatal DRD2/3 availability [22], which has also been reported in obesity, schizophrenia, and schizoaffective disorder [23]. Additionally, TaqIA allele status was found to be associated with higher DRD2/3 availability in patients with major depressive disorder and lower availability in HCs [24].
To date there have been no in vivo studies investigating DR availability and DRD2/ANKK1 TaqIA allele status in AUD, except for a postmortem study reporting a link between the binding characteristics of DRD2 in the cerebral cortex and AUD [16]. We previously showed that dorsostriatal DRD2/3 availability was reduced in AUD patients and individuals at high risk for developing AUD compared to HCs [3].
With the present study we aim to examine the DRD2/ANKK1 TaqIA allele status to investigate if the DRD2/3 availability (measured by 18F fallypride-PET) is moderated by the presence of the minor allele (A1, rs1800497 (T)) and whether this relationship depends on group affiliation (AUD versus healthy controls).
This study was part of a multicenter project investigating behavioral, genetic, and neuroimaging changes associated with reward-based learning and its relevance to the maintenance of alcohol dependence (http://www.lead-studie.de; clinical trial number NCT01679145). The study was approved by the local ethics committee (Charité–Universitätsmedizin Berlin; EA1/245/11) and all subjects gave written, informed consent before participating. Recruitment took place between 2014 and 2016. The main objective was to investigate striatal DRD2/3 availability by PET as well as glutamate level in the prefrontal cortex by magnetic resonance imaging (MRI) in patients with or at high risk of developing alcohol dependence, as well as in HCs. As we also planned to explore the neurochemical changes in AUD with potential predisposing genetic factors, blood samples were collected from the study participants for biochemical analysis and genotyping. The specific research question of the association between DRD2/ANKK1 (rs1800497) TaqI allele status and DRD2/3 availability was evaluated in a post hoc analysis. The assessed samples constituted a subsample of the above mentioned study cohort for which PET and genetic data were available.
The sample consisted of 17 HCs and 13 detoxified, abstinent patients diagnosed
with AUD. One male AUD patient had to be excluded during the process because he
showed an extremely unlikely value for receptor availability at limbic striatum
(
Subjects were matched for age, sex, education, handedness, and smoking status
(Table 1, Ref. [25, 26, 27]). AUD patients were recruited from hospitals in Berlin
and had met Diagnostic and Statistical Manual for Mental Disorders, 4th Edition
(DSM-IV) criteria for AUD for at least 3 years (American Psychiatric Association,
2000) prior to the start of the study. The patients were diagnosed at the
participating institutions by a trained clinician and their diagnosis was later
confirmed with the Composite International Diagnostic Interview (CIDI) [28, 29]
during testing. AUD patients had been abstinent for about 34.4 days and at least
72 h when they were tested, had mild withdrawal symptoms (Clinical Institute
Withdrawal Assessment for Alcohol score
Variable | Group | p value for test statistic | |
AUD (n = 12) | HC (n = 17) | ||
Demographic characteristics | |||
Sex, no. of females/males | 1/11 | 2/15 | 0.77 |
Education, years | 15.1 (3.6) | 14.8 (3.3) |
0.62 |
Age, years | 44.5 (9.3) | 45.4 (8.9) | 0.80 |
Smoker, % | 83 | 59 | 0.16 |
Clinical characteristics | |||
Duration of abstinence, days | 34.4 (23.6) | - | - |
Severity of alcohol dependence |
17.7 (5.8) | 3.5 (3.8) | |
Craving |
13.7 (8.0) | 3.8 (3.3) | |
Data are shown as mean (standard deviation) unless otherwise indicated. *Significant difference; Abbreviations: AUD, alcohol use disorder; HC, healthy control. |
All study participants with any other current substance use or dependence apart from alcohol and nicotine were excluded from the study. Healthy controls were instructed not to drink any alcoholic beverages in the 24 h before PET scanning. Compliance was verified by urine screening and breath alcohol tests. Participants with any contraindications for MRI, current pregnancy or nursing, or any neurologic or major psychiatric disorder according to DSM-IV criteria were also excluded from the study [26, 27, 31].
The study consisted of 3 separate testing sessions. In the first session, participants were asked to complete clinical questionnaires such as the Alcohol Dependence Scale or Obsessive Compulsive Drinking Scale, and underwent blood sampling for the genetic analyses MRI and PET scanning took place in the second and third sessions, respectively [25, 26].
DNA was semi-automatically extracted from whole blood with the Chemagen Magnetic Separation Module (Perkin Elmer, Boston, MA, USA). Genotyping was performed with the Infinium Psych Array Bead Chip (Illumina, San Diego, CA, USA). Three subjects (two male AUD subjects and one female healthy control) were excluded from the final sample as outliers in the genome-wide association study. We examined the status of the rs1800497 SNP, which is also known as the TaqIA polymorphism of the DRD2 gene. The minor allele is a C-to-T substitution (i.e., A1, rs1800497[T]) that is associated with a reduced number of dopamine binding sites in the brain and is thought to play a role in alcoholism, smoking, and certain neuropsychiatric disorders [16]. As there are 2 different allele types, 3 genotypes were possible (A1/A1, A1/A0, and A0/A0). Because of the very limited prevalence of A1/A1 in the HC (n = 0) and AUD (n = 1) groups, we pooled A1/A1, A1/A0, and A1 into a single group. The allele frequencies are shown in Table 2.
Associative striatum | Limbic striatum | Sensorimotor striatum | |
HC–A− (n = 7) | 24.58 (2.77) | 24.39 (1.85) | 30.75 (2.02) |
HC–A+ (n = 10) | 24.33 (3.18) | 24.56 (2.84) | 29.03 (2.55) |
AUD–A− (n = 8) |
19.74 (3.90) | 22.66 (2.06) | 25.24 (2.92) |
AUD–A+ (n = 4) | 22.56 (1.08) | 23.55 (1.61) | 27.57 (0.51) |
*Data are shown as mean (standard deviation). Abbreviations: AUD, alcohol-use disorder; A–, no risk allele; A+, at least 1 risk allele; HC, healthy control. |
For MRI, T1-weighted images were acquired using a 3 Tesla Verio scanner
(Siemens, Munich, Germany) with a magnetization prepared rapid gradient echo
(MPRAGE) sequence (isotropic resolution = 1.0 mm, repetition time = 2.3 s, echo
time = 3.03 ms, inversion time = 900 ms, and flip angle = 9
PET scanning was performed using a time-of-flight PET/computed tomography (CT) system (Gemini TF 16; Philips Medical Systems, Amsterdam, The Netherlands) [32]. Dynamic PET imaging was started simultaneously with intravenous injection of 18F-fallypride [33, 34]. Data were acquired over 4 h in 3 blocks with breaks in between. A low-dose CT scan for attenuation correction was performed before each block [35]. PET images were reconstructed using the iterative line of response–row action maximum likelihood algorithm of the scanner software with default parameter settings (3 iterations, 33 subsets, and “normal” relaxation). Head motion during the PET acquisition was corrected frame-by-frame using the realign tool of the Statistical Parametric Mapping software package [36]. Thereafter, all PET frames were coregistered to each subject’s MPRAGE MRI.
The 2-step simplified reference tissue method was used to obtain voxel-by-voxel
parametric maps of the nondisplaceable binding potential (BP
Each subject’s T1 images and BP
We used R (R 4.0.0, Bell Laboratories, Murray Hill, NJ, USA) and lme4 (Version: 1.1-25, https://cran.r-project.org/web/packages/lme4/index.html) to perform linear mixed effects analyses of the relationship
between SNP status and DRD2/3 availability across prespecified striatal regions.
We entered phenotype (AUD patients versus HCs) and SNP status (risk SNP versus no
risk SNP) into the model as fixed effects; intercepts for subjects and brain site
were entered as random effects. Visual inspection of residual plots did not
reveal any obvious deviations from homoscedasticity or normality. The p values
were obtained by likelihood ratio tests of the full model with versus without the
effect in question (maximum likelihood). DRD2/ANKK1TaqIA genotype
distribution did not significantly differ from Hardy–Weinberg equilibrium
(p
Means and standard deviations of 18F-fallypride BP
DRD2/3 Receptor Availability per site/ phenotype/ allele status.
Dopamine receptor D2 and 3 availability in the associative, sensorimotor and
limbic striatum and allele status in individuals with AUD and HC. The D2/3
availability was measured in BP
In the basic model (receptor availability ~1 + brain site +
phenotype [HC/AUD] + [1
As expected, we found a significant main effect of phenotype status (0 = HCs; 1
= AUD patients) (
The results of this study demonstrate that DRD2/3 availability is reduced in the associative, sensorimotor, and limbic striatum of AUD patients compared to HCs. This supports our hypothesis and is in line with our previous finding of reduced striatal DRD2/3 availability in AUD patients compared to individuals at high risk of AUD and HCs [3, 46]. Contrary to our hypothesis we did not find an association between the presence of the minor allele (risk allele) of the Taq1A polymorphism of the dopamine D2 receptor DRD2 gene and striatal DRD2/3 availability in AUD patients versus HCs. This result must be interpreted with caution because of the limited effect size and relatively small sample size.
Several studies have reported a link between DRD2/ANKK1 TaqIA allele status (rs1800497) and alcohol-related behaviors [16, 17, 18, 19, 20]. While DRD2/ANKK1 TaqIA allele status and striatal DRD2/3 availability have been investigated in clinical studies [22, 23, 24, 47] and postmortem samples [48, 49] ours is the first study to examine DRD2/ANKK1 TaqIA allele (rs1800497) and DRD2/3 status in AUD patients in vivo. However, a study in patients with schizophrenia found no association between DRD2 availability and TaqIA allele status by single photon emission computed tomography [50].
Despite the well-known contribution of dopaminergic impairment to AUD and other substance use disorders [3, 5, 8, 9, 51, 52], few studies have examined the relationship between DRD2/3 and genetic predisposing factors in AUD, possibly because of power issues in PET studies, which may also have contributed to the null result in this study. Therefore, our findings are noteworthy as other studies might have experienced similar limitations.
Our study population consisted of 12 AUD patients and 17 HCs, which is a typical sample size for PET studies but caused statistical difficulties when calculating allele frequencies in the same sample. Because there were 2 different allele types and hence, 3 genotypes (A1/A1, A1/A0, and A0/A0), the compared groups were relatively small. To minimize the effect of small sample size on the results and because of the lack/very small number of homozygous alleles in the HC (n = 0) and AUD (n = 1) groups, we pooled A1/A1 and A1/A0 into a single group. Many other studies have experienced the issue of insufficient data for homozygous alleles, especially in HCs [23, 50, 51, 52], and have thus been unable to detect between-group differences in the effect of DRD2/3 status for the 3 genotypes of DRD2/ANKK1 TaqIA rs1800497, as was the case in our study. Moreover, even after pooling homozygotes and heterozygotes, the compared subgroups were still relatively small, resulting in power issues in the statistical analyses. One way to overcome this limitation in the future is to pool the results of comparable PET studies with genetic data to increase the sample size.
As the molecular mechanism by which TaqIA allele status influences DRD2 expression is unknown, the functional relevance of any observed associations must be interpreted with caution. Preclinical studies have reported the regulation of ANKK1 expression by dopamine [53, 54], but no direct interaction has been demonstrated thus far.
As described above, low allele frequencies in control subjects may drive associations between DRD2 gene status and AUD [16]. However, in this study the allele frequency in HCs was comparable to that in AUD patients for the subgroup with no risk allele (A–: HC, n = 8 and AUD, n = 7), and was higher in HCs than in AUDs in the risk allele subgroup (A+: HC, n = 10 and AUD, n = 4).
Another factor potentially contributing to our results was the use of 18F-fallypride as the radioligand; this high-affinity DRD2/3 antagonist is nonselective for DRD2 and may be displaced by endogenous dopamine, although the density of DRD3 in the ventral and especially dorsal striatum is relatively low [55]. Moreover, endogenous dopamine may have influenced data acquisition although there were no stimuli during the scanning and the participants were not actively drinking at the time of the study. The use of specific DRD2 radiotracers such as [11C]NMB may improve the specificity of the results [23].
We were able to show that DRD2/3 availability is reduced in the associative, sensorimotor, and limbic striatum of AUD patients compared to HCs. This supports our previous finding of reduced striatal DRD2/3 availability in AUD patients compared to individuals at high risk of AUD and HCs [3, 46]. Contrary to our expectations we did not find any association between DRD2/ANKK1 TaqIA allele status and striatal DRD2/3 availability, although this result must be interpreted with caution because of the limited effect size and relatively small sample size.
Clarifying the association between DRD2/ANKK1 TaqIA allele status and DRD2/3 availability is important for resolving the longstanding controversy of whether dopaminergic impairment in the striatum of individuals with alcohol dependence is the result of compensatory downregulation or genetic predisposition. To this end, additional investigations that combine data from different PET studies are needed to overcome the issue of low statistical power inherent in this type of research.
The main limitation of this study was the issue of the lack of power. This problem stems from our limited sample size (N = 29) which may be appropriate for a PET study but is complicated by investigating three different genotypes in two subgroups (AUD versus HC). Nonetheless, similar studies in this research field, possess comparable sample sizes such as Eisenstein et al. (n = 39; n = 18) and Savitz et al. (N = 36) [23, 24]. Other studies provide larger samples, but investigated healthy controls only such as Pohjalainen et al. and Hirvonen et al. (both used the same sample of N = 54) [12, 51, 52]. To be able to evade this issue it is important to pool data between research groups. Therefore, the dataset of this study is available online (https://osf.io/t4brz/) to make it accessible to other researchers of this field.
GS, JG, TG, MS and EF conceived the study; GS, EF, LF, JG, AH, MS, TG, TZ and RB drafted the manuscript; GS, EF, LF, TZ, TG and RB analyzed and interpreted the data; LF created the figures; EF and JG supervised the study; TZ, AH, TG, RB, AH and MS revised the manuscript for important intellectual content. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.
The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board (or Ethics Committee) of Charité Universitätsmedizin Berlin (EA1/245/11). Written informed consent was obtained from all subjects involved in the study.
Preprint is available on https://psyarxiv.com/kx9ma; Dataset and analysis are available on https://osf.io/t4brz/; The published analysis is an additional, post hoc analysis to the preregistered trial with clinical trial number NCT01679145 available on https://clinical-trials.gov/ct2/show/NCT01679145.
We want to thank all voluntary participants for their participation in this trial. We acknowledge financial support from the Open Access Publication Fund of Charité – Universitätsmedizin Berlin and the German Research Foundation (DFG).
This research was funded by the German Research Foundation (Deutsche Forschungsgemeinschaft, FOR 1617: grants FR 3572/1-1, WA 1539/7-1, HE 2597/14-1, HE 2597/15-1, HE 2597/14 -2, HE 2597/15-2, GA 707/6-1, RA 1047/2-1, and RA 1047/2-2). Dr. med. Dr. phil. Eva Friedel is a participant in the BIH-Charité Clinician Scientist Program funded by the Charité–Universitätsmedizin Berlin and Berlin Institute of Health.
The authors declare no conflict of interest.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.